@pipeworx/tn-dmv

Tennessee Department of Safety and Homeland Security: the 186 driver service centers with their per-location service capabilities, plus the state's impaired-driving crash file. Keyless.

Part of Pipeworx — an MCP gateway connecting AI agents to 1476+ live data sources.

Why Tennessee gets its own pack

This came out of splitting a single multiplexed us-dmv tool into one pack per state agency. Tennessee's grain is a per-counter capability matrix — REAL ID, CDL skills test, handgun permit, vision test — published on the department's own ArcGIS, which has nothing in common with California's ZIP-level registration snapshot. A union schema across states leaves most arguments ignored on every call, so Tennessee's arguments are Tennessee's.

Tools

Tool	What it returns
`tn_dmv_driver_service_centers`	Address, county, hours, phone, coordinates and the 21 service capabilities each location publishes
`tn_dmv_impaired_driving_crashes`	Alcohol- and drug-involved injury and fatal crashes for 2022, grouped by age, gender, severity, agency, substance or month, with a mappable sample

tn_dmv_driver_service_centers filters: city, county, zip (prefix), name, service, limit. tn_dmv_impaired_driving_crashes filters: substance, severity, age_group, gender, agency, group_by, limit.

The distinctive fields

REAL ID, CDL skills test (only 4 of 186 locations do it), handgun permit (47 locations) and vision test are the filters worth knowing about — they are the questions a plain address list cannot answer. office_type separates the 44 full-service centers from 93 self-service kiosks, 46 county-clerk partners and 3 express counters, which matters because a kiosk performs only a handful of the listed services.

Auth

None. tnmap.tn.gov is a public state ArcGIS host.

Gotchas worth knowing

City names run together. USER_City is published as OakRidge, MtJuliet and similar, so city is matched after stripping both sides to letters and digits — a caller typing "Oak Ridge" gets the Oak Ridge center rather than a not-found.
X is longitude and Y is latitude, both as attribute columns, so no geometry round-trip is needed.
Two HTML blob columns had to be excluded. USER_Website_Name and USER_Website_Services are pre-rendered <br>-separated markup that duplicates columns already returned and roughly triples the payload; with them included the full-layer fetch failed outright rather than merely running slowly. The query names its fields explicitly to leave them behind.
The crash layer is person-level, and 2022 only. SAFETY/DUICrashesArrestsAgeGender layer 0 holds one row per person involved: ~1,543 person records across ~603 distinct crashes. Both numbers are reported (person_records, distinct_crashes) because quoting the row count as a crash count would overstate it by 2.5×. It covers injury and fatal crashes; property-damage-only crashes sit outside it. The whole file is calendar year 2022, reported as as_of.
State troopers are not called "Highway Patrol" in the data. They appear as THP District 3 - Nashville and seven other districts, so agency: "THP" selects all of them; agency: "HIGHWAY PATROL" matches nothing.
"Unknown" is a real age group, the agency's own value for records where age was never recorded, and it is the largest bucket (386 of 1,543). It is left in rather than silently dropped.
The wider SAFETY/ ArcGIS folder also carries DeerCrashes2019to2023 and I24_CrashHotspots. Both are cartographic multi-layer map services (severity-split display layers, basemap boundaries, a weighted-likelihood raster) rather than one queryable table, so they are left out; the DUI layer was the one that returned real, aggregatable records.

Data sources

Both verified live 2026-07-30.

Quick Start

Add to your MCP client (Claude Desktop, Cursor, Windsurf, etc.):

config.json

{
  "mcpServers": {
    "tn-dmv": {
      "url": "https://gateway.pipeworx.io/tn-dmv/mcp"
    }
  }
}

What this endpoint actually serves

tools/list at https://gateway.pipeworx.io/tn-dmv/mcp returns the tools in the table above plus the shared Pipeworx meta-tools — ask_pipeworx, discover_tools, search_within, remember/recall and the rest of the gateway-wide set. So the tool count you see is larger than this table: a single-pack endpoint currently lists roughly 30 shared tools alongside the pack's own. The connection's initialize response states its exact scope, and is the authoritative answer for a given day.

This is deliberate, not multiplexing by accident. The meta-tools are what let a scoped connection answer a question this pack does not cover — via ask_pipeworx, which routes across the whole catalog — without you adding a second MCP server. There is currently no way to mount a pack endpoint without them; if the extra schemas cost you more context than the routing is worth, connect to the full gateway once rather than to several pack endpoints.

Or connect to the full Pipeworx gateway to get every pack's tools listed directly, instead of just this one's:

config.json

{
  "mcpServers": {
    "pipeworx": {
      "url": "https://gateway.pipeworx.io/mcp"
    }
  }
}

Both URLs reach the same gateway and the same 1476+ data sources. The only difference is which pack's tools are listed directly; ask_pipeworx reaches all of them from either one.

Using with ask_pipeworx

Instead of calling tools directly, you can ask questions in plain English — this works on the pack endpoint above as well as on the full gateway:

Code

ask_pipeworx({ question: "your question about Tn Dmv data" })

The gateway picks the right tool and fills the arguments automatically.

License

MIT

Tennessee Driver Services README